The problem of fast transportation of a large number of passengers in large cities remains relevant at all times. Searching for the optimal combination of urban transport routes for a large city is a time-consuming task that requires an automated method of solution. The main idea of the work is to use the genetic algorithm for multi-route search, because it significantly saves memory and other computer resources in comparison with the algorithm of full selection. Thus, the genetic algorithm allows for search of multi-route combination on a much larger graph. A method of genetic algorithm adaptation for the optimal routes planning is described in the paper. Moreover, in the course of work there was an alternative algorithm for the optimal selection of a transport routes combination for transport network with a small roads number. This algorithm obtains the absolute optimal solution by evaluating all possible route options. In the work, this algorithm was used to evaluate the performance of the genetic algorithm and adjust its parameters.